How Does a Broker See Which Agents Actually Run the Process?

You are managing a team you cannot hear

Ask your agents how the listing appointment went and every one of them says it went great. Ask how the buyer consultation went and the answer is the same. Then look at the ratios. One agent converts most of their appointments. Another converts a third. Both say it went great. You have no idea what the difference is, because you were not in the room and neither was anyone else.

This is the broker's specific problem. In most sales organizations the manager can at least sit near the phones. In a brokerage the sale happens in a stranger's living room or a coffee shop, run by an independent contractor who reports to you by text. You are managing a team whose work you have never heard. 73% of sales managers spend under 30 minutes a week coaching a rep (ATD 2025, Gong 2025), and in a brokerage most of that time is spent on the CRM, not the conversation.

So you manage by proxy. Closings, pending, appointments set, CRM activity. All of it lags the actual behavior by weeks or months. By the time the ratio tells you an agent is losing listings, they have lost a season of them.

Why closings and CRM activity lie to you

Closings are the result of appointments from two or three months ago. An agent who stopped asking for the listing in March still closes deals in May from the pipeline they built in January. The number looks fine until it does not, and by then the habit is six months old. Only 29% of companies can see whether training changed what their reps do, and closings are why: they measure the outcome, not the behavior.

CRM activity is worse. It measures whether the agent logged something, not what they said. An agent can log twenty follow-up calls a week and every one of them can be a two-minute check-in that moves nothing. Another agent logs five and each one lands a specific reason for the seller to call back. The CRM cannot tell them apart. Neither can you, from the dashboard.

What you need is the behavior itself. Did the agent ask before presenting? Did they raise the other agents first? Did they hold the fee? Did they ask for the listing? Did they ask for the buyer agreement? Did they book the next appointment before leaving? Those are yes or no questions about a conversation, and the only way to answer them is to hear it.

Hear the appointment, on both sides

With the seller's or buyer's consent stated at the start, the agent records the appointment from their phone. Zoom consultations are recorded the same way. Every recording is transcribed and scored on six skills: value stacking, objection handling, close timing, price presentation, follow-up setup, and talk-to-listen ratio. Then it is scored against your brokerage's own process, which carries 60% of the weight, because your steps matter more than a generic method.

Your process is the part most brokers skip, and it is the part that makes the score yours. Write down what a listing appointment at your brokerage has to include, step by step. Write down the buyer consultation the same way, including the agreement ask. Those steps become the standard. When an agent skips the agreement conversation, the score says so, and the agent cannot argue with a standard they were given in writing.

The best and worst moment of every appointment are timestamped. You do not listen to 45 minutes. You open the appointment that scored 58, jump to minute 31, and hear the seller ask about the fee and your agent fold. That is the coaching conversation, and it takes five minutes because it starts from a fact.

• Value stacking
• Objection handling
• Close timing
• Price presentation
• Follow-up setup
• Talk-to-listen ratio
• Your brokerage's own process at 60% weight

What a low score should trigger

A score with no consequence is a report card nobody reads. The value is in what happens next. A strong appointment becomes an example for the team. A weak one becomes practice on the exact moment that failed, quickly, while the agent still remembers the room.

In Revenue OS, an appointment under 80 becomes a redo due 6pm the next business day, run against an AI seller or buyer seeded with the real customer's words from the transcript. The agent who folded on the fee at minute 31 hears that seller's sentence again tomorrow and has to hold it this time. Managers get a coaching queue sorted by urgency, so the 30 minutes a week go to the agent and the moment that need it most.

This also fixes the consent-and-trust problem agents raise. The recording is not surveillance. It is the thing that gets them a specific fix instead of a vague talk about their ratio. Most agents, once they have heard their own worst moment and fixed it, stop objecting to the recording.

One number per agent, every night

Hearing appointments tells you what happened in the room. It does not tell you whether the agent is doing the daily work that improves the room. For that you need a behavior score. One number per agent, 0 to 100, recalculated nightly, weighted 30% training completion, 25% daily tasks, 25% role play quality, and 20% manager enforcement.

The weights are deliberate. Training and practice are more than half, because those are the behaviors independent agents skip first when nobody is watching. Manager enforcement is in there so that a team lead who lets low scores slide shows up in the pattern too. The score is visible to every agent, ranked. Nobody wants to be last on a list their peers can see, and the agents at the top want the proof.

Your dashboard becomes simple. Agents ranked. A coaching queue by urgency. The objections your team loses most, across both sides of the business. Follow-up outcomes per agent: responded, booked, closed, no response. You see who is running appointments and who is running open houses, and you see it tonight instead of at the end of the quarter.

What this looks like on a Tuesday

Morning. You open one screen. Twelve agents, twelve numbers, sorted. Two are below 70. One skipped the lesson three days running. The other is practicing but scoring under 60 on the buyer consultation role play, which means the agreement conversation is not there yet. Two different problems, two different ten-minute conversations, and you knew which was which before you said hello.

Then the coaching queue. Three appointments from yesterday scored under 80. You open the worst one, jump to the timestamp, hear the moment, and confirm the redo is due at six. Twenty minutes total. You did not ride along. You did not ask anyone how it went. You heard it, scored it, and acted on it, and every agent on the team knows the same thing happens tomorrow. That is how a broker who cannot be in the room still knows what happened in it, and it is the foundation a team needs to break its revenue record.